From ea453f819078893ef7d6dc5b2faf1f543919d77e Mon Sep 17 00:00:00 2001
From: Zac Medico <zmedico@gentoo.org>
Date: Fri, 2 Nov 2007 05:29:03 +0000
Subject: Rewrite the dblink.getcontents() code to use str.split(" ") for
 splitting CONTENTS lines so that even file paths that end with spaces can be
 handled. This patch makes the fix for bug #196836#c6 more complete. Some code
 for parsing old malformed symlink entries has been removed sinces it's
 probably not useful or worth maintaining anymore. (trunk r8337)

svn path=/main/branches/2.1.2/; revision=8364
---
 pym/portage.py | 71 +++++++++++++++++++++++++++++++++++++++-------------------
 1 file changed, 48 insertions(+), 23 deletions(-)

diff --git a/pym/portage.py b/pym/portage.py
index 5e3cb6bd6..af27cf19d 100644
--- a/pym/portage.py
+++ b/pym/portage.py
@@ -7185,6 +7185,13 @@ class dblink:
 
 	import re
 	_normalize_needed = re.compile(r'.*//.*|^[^/]|.+/$|(^|.*/)\.\.?(/.*|$)')
+	_contents_split_counts = {
+		"dev": 2,
+		"dir": 2,
+		"fif": 2,
+		"obj": 4,
+		"sym": 5
+	}
 
 	def __init__(self, cat, pkg, myroot, mysettings, treetype=None,
 		vartree=None):
@@ -7320,6 +7327,7 @@ class dblink:
 		myc.close()
 		null_byte = "\0"
 		normalize_needed = self._normalize_needed
+		contents_split_counts = self._contents_split_counts
 		myroot = self.myroot
 		if myroot == os.path.sep:
 			myroot = None
@@ -7332,7 +7340,37 @@ class dblink:
 					"file, line %d: '%s'\n" % (pos, contents_file),
 					noiselevel=-1)
 				continue
-			mydat = line.split()
+			line = line.rstrip("\n")
+			# Split on " " so that even file paths that
+			# end with spaces can be handled.
+			mydat = line.split(" ")
+			entry_type = mydat[0] # empty string if line is empty
+			correct_split_count = contents_split_counts.get(entry_type)
+			if correct_split_count and len(mydat) > correct_split_count:
+				# Apparently file paths contain spaces, so reassemble
+				# the split have the correct_split_count.
+				newsplit = [entry_type]
+				spaces_total = len(mydat) - correct_split_count
+				if entry_type == "sym":
+					try:
+						splitter = mydat.index("->", 2, len(mydat) - 2)
+					except ValueError:
+						writemsg("!!! Unrecognized CONTENTS entry on " + \
+							"line %d: '%s'\n" % (pos, line), noiselevel=-1)
+						continue
+					spaces_in_path = splitter - 2
+					spaces_in_target = spaces_total - spaces_in_path
+					newsplit.append(" ".join(mydat[1:splitter]))
+					newsplit.append("->")
+					target_end = splitter + spaces_in_target + 2
+					newsplit.append(" ".join(mydat[splitter + 1:target_end]))
+					newsplit.extend(mydat[target_end:])
+				else:
+					path_end = spaces_total + 2
+					newsplit.append(" ".join(mydat[1:path_end]))
+					newsplit.extend(mydat[path_end:])
+				mydat = newsplit
+
 			# we do this so we can remove from non-root filesystems
 			# (use the ROOT var to allow maintenance on other partitions)
 			try:
@@ -7342,34 +7380,21 @@ class dblink:
 						mydat[1] = os.path.sep + mydat[1]
 				if myroot:
 					mydat[1] = os.path.join(myroot, mydat[1].lstrip(os.path.sep))
-				if mydat[0]=="obj":
+				if mydat[0] == "obj":
 					#format: type, mtime, md5sum
-					pkgfiles[" ".join(mydat[1:-2])]=[mydat[0], mydat[-1], mydat[-2]]
-				elif mydat[0]=="dir":
+					pkgfiles[mydat[1]] = [mydat[0], mydat[3], mydat[2]]
+				elif mydat[0] == "dir":
 					#format: type
-					pkgfiles[" ".join(mydat[1:])]=[mydat[0] ]
-				elif mydat[0]=="sym":
+					pkgfiles[mydat[1]] = [mydat[0]]
+				elif mydat[0] == "sym":
 					#format: type, mtime, dest
-					x=len(mydat)-1
-					if (x >= 13) and (mydat[-1][-1]==')'): # Old/Broken symlink entry
-						mydat = mydat[:-10]+[mydat[-10:][stat.ST_MTIME][:-1]]
-						writemsg("FIXED SYMLINK LINE: %s\n" % mydat, 1)
-						x=len(mydat)-1
-					splitter=-1
-					while(x>=0):
-						if mydat[x]=="->":
-							splitter=x
-							break
-						x=x-1
-					if splitter==-1:
-						return None
-					pkgfiles[" ".join(mydat[1:splitter])]=[mydat[0], mydat[-1], " ".join(mydat[(splitter+1):-1])]
-				elif mydat[0]=="dev":
+					pkgfiles[mydat[1]] = [mydat[0], mydat[4], mydat[3]]
+				elif mydat[0] == "dev":
 					#format: type
-					pkgfiles[" ".join(mydat[1:])]=[mydat[0] ]
+					pkgfiles[mydat[1]] = [mydat[0]]
 				elif mydat[0]=="fif":
 					#format: type
-					pkgfiles[" ".join(mydat[1:])]=[mydat[0]]
+					pkgfiles[mydat[1]] = [mydat[0]]
 				else:
 					writemsg("!!! Unrecognized CONTENTS entry on " + \
 						"line %d: '%s'\n" % (pos, line), noiselevel=-1)
-- 
cgit v1.2.3-1-g7c22